Carbon nanotubes are a material having a structure in which graphene sheets are rolled into a cylindrical shape and having a one-dimensional structure having a very large aspect ratio (see Non Patent Literature 1). The carbon nanotubes are known to have mechanically excellent strength and flexibility, semiconducting and metallic conductivity, and further, chemically very stable properties. For methods for producing carbon nanotubes, an arc discharge method, a laser vaporization method, a chemical vapor deposition method (hereinafter, referred to as a CVD method), and the like are reported. Particularly, the CVD method is a synthesis method that receives attention as a synthesis method suitable for mass production, continuous operation, and higher purity (see “Basics and Applications of Carbon Nanotubes” jointly edited by Riichiro Saito and Hisanori Shinohara, BAIFUKAN, published in 2004).
Particularly, single-walled carbon nanotubes (hereinafter, referred to as “SWCNTs”) have been confirmed to exhibit metallic properties or semiconducting properties, depending on the way of rolling and their diameter, and applications to electrical and electronic devices and the like have been expected. For the synthesis of SWCNTs, a catalytic CVD method in which nanotubes are grown (for example, see Non Patent Literature 2) has become a mainstream. This catalytic CVD method uses nanoparticles of metal as a catalyst. And, while a carbon source which is a gas is fed, the carbon source is pyrolyzed at high temperature to grow nanotubes from the nanoparticles of metal, the catalyst. At this time, the nanotubes are produced using the catalyst, which is the nanoparticles, in a gas phase-dispersed state (an A method). In addition, there is also a method using the catalyst, which is the nanoparticles, in a substrate-supported state (a B method). The A method and the B method each have advantages and disadvantages.
[Regarding Existing SWCNT Production Methods]
The outline of the A method of the gas phase-dispersed catalyst is illustrated in FIG. 14. In this method, a catalyst source and a carbon source are simultaneously fed into an externally heated reactor to perform the synthesis of nanotubes. Examples of typical synthesis methods classified into this A method include a HiPco method (for example, see Non Patent Literature 3). This A method can effectively use the three-dimensional space of the reactor. But, since the catalyst is entrained in a reaction gas, time that the catalyst remains in the reactor is short, and the catalyst is mixed into the nanotubes, a product. In addition, since the nanoparticles of the catalyst are as small as several nm, and aggregation is fast, it is difficult to increase the spatial concentration of the catalyst, and nanotube productivity per L of reactor volume is about 1 g/day.
The outline of the B method of the substrate-supported catalyst is illustrated in FIG. 15. In this B method, the catalyst is supported on a substrate, and a carbon source is fed onto the catalyst to grow nanotubes on the catalyst. Super Growth method (for example, see Non Patent Literature 4) and the like are classified as this B method, and its typical synthesis methods. In this B method, fast nanotube growth is possible. For example, fast growth at 2.5 mm/10 min is performed (Non Patent Literature 4). In addition, the catalyst is fixed on the substrate, and thus, the catalyst is prevented from being mixed into the synthesized nanotubes. But, since in the reactor, only a two-dimensional space which is a plane can be used, space use in the reactor is poor, compared with the A method.
Further, in the B method, a separation step for the separation of the synthesized nanotubes is necessary. In the case of the mass production of nanotubes, the repeated use of a substrate with a catalyst is indispensable, and this technique has not been established yet. There are many patent literatures in which carbon nanotubes are synthesized with a fluidized bed by the B method, using particles, instead of the substrate, for the fixing of the catalyst. For example, in Patent Literature 1, an apparatus for producing a tubular carbon substance is disclosed. Here, a fluidized-bed reaction furnace in which carbon nanotubes are continuously produced is disclosed (see the paragraph [0007] of Patent Literature 1).
Further, examples of techniques for producing carbon nanotubes, using a fluidized bed, include a CoMoCAT (registered trademark) production method (URL: http://www.ou.edu/engineering/nanotube/comocat.html). This production technique is a method of contacting a catalyst containing a group VIII metal, such as cobalt (Co), or a group VIa metal, such as molybdenum (Mo), with a carbon-containing gas to produce carbon nanotubes, and has been developed by the University of Oklahoma in the United States, and put to practical use by SouthWest NanoTechnologies Inc. Patent Literatures 2 to 10 are U.S. Patents regarding this technique for producing carbon nanotubes, a list of patents that the University of Oklahoma in the United States possesses.
In these synthesis methods with a fluidized bed, a catalyst is supported on support particles of porous silica or the like to synthesize nanotubes, the nanotubes are removed together with the support particles from a fluidized-bed apparatus, and the support particles and the catalyst are dissolved with an acid or the like to recover the nanotubes. But, the support particles with catalyst particles are used only once and then thrown away, the step of removing the support and the catalyst from the nanotubes is complicated, and operation is batch-wise and productivity is not high, and therefore, the price of SWCNTs is 50000 yen/g or more and is very expensive.
In addition, in recent years, demand for hydrogen (H2) as clean energy has been increasing. Therefore, methods for efficiently producing hydrogen have been studied. As conventional methods for producing hydrogen, a method of producing hydrogen by a steam reforming reaction, using hydrocarbon as a source, is common (for example, see Patent Literatures 11 and 12).